Adults' understanding of the mathematical and science knowledge underlying nutrition information was studied using a low cost HyperCard package. This package, which runs on a Powerbook, allows the users to access information in a nutrition database in a nonlinear hypertext fashion using an alphabetical index, selecting 'hot' text or clicking on various navigation tools. The users' paths through the package were captured using HyperTalk scripts and analysed to shed light on user interface issues and the underlying cognitive processes. Some users encountered disorientation, and many used only the index system to retrieve information. This study contributes to the continual search for guidelines to improve user interface in hypertext systems for real life applications.
In daily life adults are constantly exposed to a range of alternative and often conflicting information and advice about nutrition. The adults themselves have their own working knowledge about food, derived mainly from family or traditional practices. With the current concerns about improving health in the community, a common view to get changes in nutritional practices is 'to tell people to do it' (Clements, 1986, p. 229). This strategy may miss the point because adults cannot be conceived as empty vessels into which knowledge is poured. A constructivist perspective suggests that adults actively transform the given information in the light of their previous experiences and knowledge so that it can guide actions in particular situations. Thus it is important to understand in detail these prior conceptions of nutrition, which this study attempts to delve into. Further discussion of this theoretical framework is presented elsewhere (Schibeci & Wong, 1993). We believe that it is timely, in Australia, to develop strategies which will provide a suitable constructivist vehicle for nutrition education and an interactive hypertext package seems to provide a potentially powerful tool to explore this shift. The remaining part of this paper describes the development and testing of the HyperNutrition package.
The study commenced in April 1993 with a group of 25 university students. During this phase of the study, the package and its use was revised three times in the light of the responses from the users. The second phase involved 19 women who were members of a Yoga class or the Nursing Mothers' Association. All were volunteers. They were interviewed and they generally spent between half to one hour working with the package.
Figure 1: The help card
Figure 2 shows the card about the statement, "Sugar is bad for your health." The respondent is asked to give two responses to this statement. The first response is given without consulting the Dictionary; this reflects the respondent's initial opinion about the statement. After entering the first response, the respondent is asked to browse through the Dictionary to look for relevant information. After the search, the respondent returns to the same statement card and inputs the second response. Comparing these two responses to the same statement may indicate a change in conceptions. However, given the short time available for each interview, we do not expect major changes in opinions to occur. Instead, we are interested primarily in testing the usefulness of this methodology and in identifying any other information mentioned by the respondents that is missing from the Dictionary. This enables us to decide what additional information to include in the Dictionary.
Figure 2: The card about sugar
Not all questionnaire items require two responses, however. The other items ask the respondents to work out their BMI (Body Mass Index), to compute the energy values of foods, to comment on food labels, and to enter their biographical data. There is also a Convert card, useful for people who need to convert between different units of measurement, such as feet to metres, calories to joules, and so on (see figure 3).
Figure 3: The conversion card
From the Survey stack, the user can open the Dictionary by clicking on the Dictionary icon or any text in the Survey Stack. The Dictionary stack occupies a smaller window (see figure 4) so that the user can move easily from the Dictionary window back to the Survey stack by clicking on the Survey window.
At the end of each interview, the time spent on each card and the answers typed into the answer boxes were exported as text files. Figure 5 shows part of the record obtained from Mary (not her real name). To answer the statement about sugar, she had searched four cards in the Dictionary: Energy sources, Joule, Recommended dietary allowances, and Sugar.
Figure 4: The survey window and the dictionary window
In each case, she started her search from the Index card. After the search, she changed her opinion about the Sugar statement from 'disagree' to 'agree'. Her reasons for the change were recorded on audio tape.
4/22/93 - 12:36
SI-Title Card -89sec
S11-Sugar-D1-card id 8011-27sec
Figure 5: Part of the recorded data from Mary
During the trials with university students, some of them became lost when they moved forwards and backwards from the Survey stack to the Dictionary stack. They also complained about not being able to remember what the questionnaire items wanted when they were searching for information in the Dictionary. While disorientation is fairly common in complicated hypertext systems, it surprises us to find this a problem with tertiary students working on a simple package such as ours. Getting the respondents to type in their responses on the computer and asking them to respond to the same statement twice were too time consuming. As a result of this trial, we simplified the questionnaire, and for the study involving the two community groups, we used a printed survey instead of the electronic one. This saved time and reduced the cognitive demand on the respondents. However, the disadvantages are that some information has to be typed in later, instead of being extracted as text files, and that we could not readily assess any changes in conceptions.
Figure 6: The index card of the dictionary
By confronting our respondents with different and sometime conflicting views about the same topic, we wish to study what they make sense of thew views, thereby displaying an indication of their thinking processes. A typical card, in this case the food plan recommended by the Heart Foundation, is shown in figure 7 . In addition to the nutrition information, each card also displays its name, its number in relation to the total number of cards (which indicates the size of the stack), the navigation buttons, and a field for entering users' comments.
Figure 7: Food plan
We learned from the first phase of our study that many respondents queried the reliability of the information in the Dictionary. In particular, they wished to know the source of the information since they tended to trust information that was taken from authoritative sources such as the Heart Foundation than from popular magazines. Thus we have added an 'Authors' menu item to allow the users to look up the sources if they wish to. For example, if they look up 'CHOICE' in the Author menu, they will get the information shown in figure 8.
Figure 8: Information about CHOICE
When a Powerbook 170 became available, we included a digitised voice recording to give instructions on how to find information in the Dictionary. The instruction is played when the Index card is opened for the first time. Subsequently the voice instruction can he repeated by clicking on the Play button. Our respondents, in fact, did not pay much attention to this recorded message; they kept asking the interviewer for the instructions! This seems to support Norman's (1993) claim that people have trouble remembering alternatives given in voice messages; besides, listening is slower than reading the same instructions on the screen.
Figure 9: Dialogue box with Find
At the end of each interview, the respondents rated on a checklist about their use of the stack. On the whole, they agreed strongly that Dictionary was easy to use and contained useful information about nutrition.
D 1 g-exercise and nutrition-24sec
Adults who have limited computer experience had difficulty remembering what questions they were supposed to answer when they used two stacks. Using a single stack with a printed questionnaire alleviates part of this disorientation, but this problem will become serious when the Dictionary is further expended. A graphical browser to provide an overview of the content and structure of the database will be considered (Smith & Wilson, 1993).
Our respondents tended to use a limited search strategy, namely the Index-to-card system found in conventional books. We agree with Wright and Lickorish (1990) that a combination of information retrieval strategies will be suitable in using hypertext systems to serve different tasks. In future study the users will be given some training on how to use various strategies, in particular the Find button and click text methods, to retrieve information from the Dictionary.
The prevailing method to inform people how much foods to take is to present the information in numerical, tabular, verbal, or pictorial form. More recently some nutritionists have suggested that presenting the actual amounts of the foods on a plate might be a more effective strategy. Using real foods is not very convenient in research. Instead, video pictures of the foods captured as QuickTime movies may serve the same visual functions as the real foods. It is worth exploring which representational forms are easiest to understand and to use to plan food intake.
A lot of effort has been expended to input a large amount of nutrition information from various sources into the Dictionary. However, in the present study, the respondents had time to use only part of this database. Our next strategy is to let each respondent have the Powerbook perhaps for a week so that he/she can browse through the Dictionary at leisure according to individual needs and interests. To facilitate this use, a bookmark will be included so that the user knows which cards have been visited and can return immediately to the place lasted visited. Interviews will be conducted at the beginning and end of this extended period of use. This approach will contribute to field studies that "address the useability of hypertext in the users' own environment rather than in a laboratory setting" (Nielsen & Lyngbaek, 1990, p.64). A low cost package is particularly appropriate for this extended study. A similar approach could also be used to understand how school students understand about nutrition.
Other cosmetic elements such as voice, colour, graphics, and animations will be examined for their usefulness to enhance the visual and audio impacts of the package. The portability of the stack to the DOS or Windows environment will also be explored.
To summarise, we have described how the stacks were developed to serve the needs of our research to study adults' cognition about nutrition and to suit the ability of users who have limited exposure to multimedia technology. Our challenge is produce a product that is useful to health educators, school students and the general public. Further testings with different target groups are being planned.
Driver, R., Guesne, E. & Tiberghien, A. (1985). Children's ideas in science. Milton Keynes: Open University.
Malone, L. A. & Taylor, P. C. S. (Eds) (1993). Constructivist interpretations of teaching and learning mathematics. Perth: Curtin University of Technology.
Nielsen, J. & Lyngbaek, U. (1990). Two field studies of hypermedia useability. In R. McAleese & C. Green (Eds), Hypertext: State of the art (pp. 64-72). Oxford: Intellect.
Norman, D. A. (1993). Things that make us smart: Defending human attributes in the age of the machine. Reading, Mass: Addison-Wesley.
Schibeci, R. & Wong, K. Y. (1993, November). Transforming knowledge: Adults' understanding of human nutrition. Paper presented at the Annual Conference of Australian Association for Research in Education, Perth.
Smith, P. A. & Wilson, J. R. (1993). Navigation in hypertext through virtual environments. Applied Ergonomics, 24(4), 271-278.
Wright, P. & Lickorish, A. (1990). An empirical comparison of two navigation systems for two hypertexts. In R. McAleese & C. Green (Eds), Hypertext: State of the art (pp. 84-93). Oxford: Intellect.
Wynne, B. (1991). Knowledges in context. Science, Technology and Human Values, 16(1), 111-121.
Acknowledgments: We are grateful to Murdoch University for funding this research in 1993. We also wish to thank Linda McGluckin for conducting the interviews and preparing the transcripts.
|Authors: Khoon Yoong Wong|
Lecturer in Distance Education, External Studies Unit
Murdoch University, Murdoch WA 6150 Tel. 09 360 6015, Fax. 09 310 4929
Please cite as: Wong, K. Y. and Schibeci, R. (1994). Research on adults' understanding of nutrition knowledge using a low cost multimedia package in HyperCard. In C. McBeath and R. Atkinson (Eds), Proceedings of the Second International Interactive Multimedia Symposium, 589-593. Perth, Western Australia, 23-28 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1994/qz/wong.html